I’m no expert in micro-economics or psychophysics, but:

I’d guess that it depends on the relationship between the raw variable and utility.
I can see how distance travelled for Bangladeshis could be roughly linearly related to time and that the utility of time saved would roughly be linear to the amount of time saved.

Likewise, presumably human perceptual systems are designed to logarithmically transform raw stimuli (e.g., sound waves, light intensity) because the adaptive utility of distinctions operate logarithmically (i.e., little differences at low levels are important; only big differences at high levels are important).

It’s also interesting to think about situations where the transformation is not adaptive.